This invention pertains to a method and device to use certain enhancer molecules tethered to an artificial surface in such a manner that the molecules retain their quaternary structure and that the surface produced promotes normal corneal epithelialization.
The Corneal Epithelium
The cornea serves several important functions for vision, e.g., refracting  greater than 80% of the incoming light onto the retina, filtering out harmful UV rays, and maintaining an optical xe2x80x9cwindow.xe2x80x9d The cornea is composed of five structural layers: the epithelium, Bowman""s layer, stroma, Descemet""s membrane, and endothelium. The outermost layer, the corneal epithelium, is a multi-layered structure with a complex arrangement of intercellular junctions, chemical signaling, and nerve endings. Similar to other epithelial layers throughout the body, the corneal epithelium prevents entry of pathogens, provides a barrier against fluid loss, and protects against abrasive wounding. The epithelium is separated from the external environment only by a layer of fluid, the tears.
The corneal epithelium is composed of three types of cellsxe2x80x94the basal cells (1 layer), wing cells (1-3 layers), and squamous cells (3-4 layers)xe2x80x94which adhere to one another by tight cell junctions. The basal cells also form strong adhesion complexes with the underlying extracellular matrix and ultimately with Bowman""s layer. Bowman""s layer, the anterior-most layer of the corneal stroma, is an acellular zone consisting of collagen fibrils and associated proteoglycans which are densely woven in a random fashion into a felt-like matrix. Of the cells in the epithelium, only the basal cells have mitotic capabilities. Like all stratified epithelia in the body, the corneal epithelium is self-renewing; complete cellular turnover occurs every 5-7 days. Generally, after the basal cells undergo mitosis, the daughter cells begin to move outward toward terminal differentiation and eventual desquamation. See C. Hanna et al., xe2x80x9cCell production and migration in the epithelial layer of the cornea,xe2x80x9d Arch. Ophthalmol., vol. 64, pp. 536 (1960).
Epithelial Cell Migration During Wound Healing
Natural corneal epithelialization occurs in response to a wound. Epithelial recovery following an external injury is a complex process characterized by two phases, the latent phase and the healing phase. See C. E. Crosson et al., xe2x80x9cEpithelial wound closure in the rabbit cornea: A biphasic process,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 21, pp. 464-473 (1986). These phases are thought to be initiated by the immediate expression of a nucleoprotein-encoding protooncogene (c-fos). Immediately after full-thickness wounding, the damaged cells adjacent to the wound edge lose surface microvilli. Polymorphonuclear neutrophils (PMNs) arrive via the tear film and begin the process of debriding cellular remnants. During the first six hours of epithelial debridement (latent phase), a single layer of epithelial cells at the wound margin becomes motile by forming cellular processes at the wound edge and releasing their hemidesmosomal attachments to the basement membrane. Fibronectin, the glycoprotein widely involved in cell-to-cell and cell-to-substrate interactions, is deposited from the tears onto the denuded corneal surface, along with fibrinogen and fibrin. See L. S. Fujiikawa et al., xe2x80x9cFibronectin in healing rabbit cornea wounds,xe2x80x9d Lab. Invest., vol. 45, pp. 120 (1981).
The initiation of lateral migration of epithelial cells over the wound signifies the onset of the healing phase. In association with the proteins fodrin and vinculin, the assembly and disassembly of intracellular actin filaments provides cytoskeletal locomotive support during this pre-mitotic stage of the healing phase. Adhesion of the migrating epithelial monolayer to the stroma is thought to be mediated by the glycoprotein fibronectin. See S. Hiraki et al., xe2x80x9cBiochemical and histological findings on the effect of fibronectin in rabbits with experimental corneal disorders,xe2x80x9d Arzneim-Forch/Drug Res., vol. 40, pp. 1336-1340 (1990). Fibronectin contains both cell-specific binding sequences and a binding region for heparin sulfate and type IV collagen (basement membrane components). Fibronectin is thought to provide a temporary subepithelial matrix on which the epithelial cells can migrate, in repetitive cycles during which the cells cleave their attachments, advance, and then form new attachments. Plasmin, generated by urokinase-like plasminogen activator (uPA), is thought to cleave the fibronectin at the basal cell surface to release the leading edge of the epithelium. See M. Berman, xe2x80x9cThe pathogenesis of corneal epithelial defects,xe2x80x9d In Healing Processes in the Cornea, (Beuerman et al. eds), Portfolio Publishing Company, The Woodlands, Tex. (1989). After about 24 hours, the migrating epithelial cells begin to proliferate, and the epithelium attaches to the basement membrane more firmly via newly synthesized hemidesmosomes and associated type VII collagen containing anchoring filaments. See I. K. Gipson et al., xe2x80x9cAnchoring fibrils form a complex network in human and rabbit cornea,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 28, pp. 212-220 (1987). The anchoring filaments pass through the basement membrane and are contiguous with anchoring fibrils that terminate as anchoring plaques in Bowman""s layer. See I. K. Gipson et al., 1987; and T. Nishida et al., xe2x80x9cExpression of fibronectin receptors in corneal epithelial cells,xe2x80x9d In: Healing Processes in the Cornea (Beuerman R W, Crosson C E, Kaufman H E (eds)), Portfolio Publishing Company, The Woodlands, Tex. (1989). The proteins laminin, K-laminin, talin, integrin, and kalinin also play roles in the attachment of the epithelium to the stroma. See L. M. Sorokin et al., xe2x80x9cDevelopmental regulation of the laminin xcex15 chain suggests a role in epithelial and endothelial cell maturation,xe2x80x9d Dev. Biol., vol. 189, pp. 285-300(1997); L. S. Grushkin-lerner et al., xe2x80x9cExpression of integrin receptors on plasma membranes of primary corneal epithelial cells is matrix specific,xe2x80x9d Exp. Eye Res. vol. 64, pp. 323-334 (1997); M. P. Marinkovich et al., xe2x80x9cThe dermal-epidermal junction of human skin contains a novel laminin variant,xe2x80x9d J. Cell Biol., vol. 119, pp. 696-703 (1992); A. Horowitz et al., xe2x80x9cInteraction of plasma membrane fibronectin receptor with talin: a transmembrane linkage,xe2x80x9d Nature, vol. 320, pp. 531-533 (1986); T. Paallysaho et al., xe2x80x9cEpithelial cell-substrate adhesion in the cornea: Localization of actin, talin, integrin and fibronectin,xe2x80x9d Exp. Eye Res., vol. 52, pp. 261-267 (1991); and P. Rousselle et al., xe2x80x9cAn epithelium-specific basement membrane adhesion molecule that is a component of anchoring filaments,xe2x80x9d J. Cell Biol., vol. 114, pp. 567-576 (1991).
Although much is known about the factors that make up the adhesion complexes, the precise sequence by which these adhesion complex components are assembled is still unknown. Certain growth factors, such as EGF, FGF, and TGF-xcex2, enhance the rate of epithelial wound healing, and human epithelial growth factor (EGF) has been specifically shown to induce a dose-dependent increase in epithelial replication in the epithelial stem cells of the corneoscleral limbus. See M. Grant et al., xe2x80x9cEffects of epidermal growth factor, and transforming growth factor-xcex2 on corneal cell chemotaxis,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 33, pp. 3292-3301 (1992); S. E. Wilson et al., xe2x80x9cFibroblast growth factor-1 receptor messenger RNA expression in corneal cells,xe2x80x9d Cornea, vol. 12, pp. 249-254(1993); and S. E. Wilson et al., xe2x80x9cEGF, basic FGF and TGF beta-1 messenger RNA production in rabbit corneal epithelial cells,xe2x80x9d Invest. Opthalmol. Vis. Sci., vol. 33, pp. 1987-1995 (1992).
Both laminin and fibronectin are multifunctional extracellular matrix proteins that play a central role in cell adhesion and migration. Expression of integrin receptors for both of these proteins has been shown to occur within hours of the ligands being detected in the matrix. See L. S. Grushkin-Lerner et al., xe2x80x9cExpression of integrin receptors on plasma membranes of primary corneal epithelial cells is matrix specific,xe2x80x9d Exp. Eye Res., vol. 64, pp. 323-334 (1997). Laminin is a major component of the intact basal lamina and has been shown to enhance epithelial cell adhesion and spreading on surfaces. See V. Trinkaus-Randall et al., xe2x80x9cModification of polymers for synthesis by corneal epithelial cells,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 32(suppl), pp. 1072 (1991). In the cornea, fibronectin is seen on the surface of healing epithelial wounds. See L. S. Fujikawa et al., xe2x80x9cFibronectin in healing rabbit corneal wounds,xe2x80x9d Lab. Invest., vol. 45, p. 120 (1981). Exogenous fibronectin has been shown to promote the healing of corneal epithelial wounds in vivo and in vitro and the healing of persistent corneal epithelial defects in humans. See T. Nishida et al., xe2x80x9cFibronectin enhancement of corneal epithelial wound healing of rabbits in vivo,xe2x80x9d Arch. Ophthalmol., vol. 102, pp. 455-456 (1984); and K. Watanabe et al., xe2x80x9cStimulatory effects of fibronectin and EGF on migration of corneal epithelial cells,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 28, p. 205 (1987).
Substance P is a corneal neurotransmitter. The cornea is heavily innervated with sensory nerve fibers. This innervation plays an important role in the maintenance of the normal structure and functions of the cornea and in the wound healing process. In clinical settings, patients who suffer from delayed epithelial wound healing or persistent corneal epithelial defects have decreased corneal sensitivity, which suggests impairment of the sensory nerve fibers. Recently a synergistic effect of substance P and IGF-1 on the stimulation of rabbit corneal epithelial cell migration and attachment to extracellular matrix proteins has been reported. See T. Nishida et al., xe2x80x9cSynergistic effects of substance P with insulin-like growth factor-1 on epithelial migration of the cornea,xe2x80x9d J. Cell Physiol., vol. 169, pp. 159-166 (1996).
If the epithelial basement membrane has not been damaged by the injury, the recovering epithelial cells are able to use it for tight adhesion to the underlying stroma, as evidenced by the reappearance of hemidesmosomes two days after injury and by tight adhesion shortly thereafter. See A. A. Khodadoust et al., xe2x80x9cAdhesion of regenerating corneal epithelium: the role of basement membrane,xe2x80x9d Am. J. Ophthalmol., vol. 65, pp. 339-348 (1968). However, if the basement membrane has been damaged, new basement membrane complexes do not begin to form until five to seven days after injury. Depending on the severity of the basement membrane damage, complete basement membrane reconstruction may take as long as nine weeks.
Recently, investigators have hypothesized that if the epithelial injury penetrates the basement membrane, such as in refractive corneal surgery, programmed cell death (apoptosis) in the underlying anterior stromal keratocytes is stimulated. Apoptosis of the keratocytes immediately after a deep epithelial injury may be mediated by modulators, such as interleukin-1 alpha and beta, that are released from injured or dead corneal epithelial cells. See S. E. Wilson et al, xe2x80x9cEpithelial injury induces keratocyte apoptosis: Hypothesized role for the interleukin-1 system in the modulation of corneal tissue organization and wound healing,xe2x80x9d Exp. Eye Res., vol. 62, pp. 325-337 (1996). The subsequent stromal wound healing response is not limited to production of collagen and other components of stromal remodeling by activated keratocytes as they repopulate the anterior corneal stroma. Activated keratocytes also secrete hepatocyte growth factor (HGF) and keratinocyte growth factor (KGF), the effects of which are not restricted to the stromal cells. These growth factors also stimulate epithelial cell proliferation and inhibit epithelial differentiation, effects that could promote epithelial hyperplasia and ultimately prolong the entire wound healing response. See S. E. Wilson et al., xe2x80x9cEffect of epidermal growth factor, hepatocyte growth factor, and keratinocyte growth factor, on proliferation, motility and differentiation of human corneal epithelial cells,xe2x80x9d Exp. Eye Res., vol. 59, pp. 665-678 (1994).
Corneal Augmentation and Replacement
Complete synthetic replacement of diseased or opacified corneas has been investigated for more than 200 years. While initial replacement studies included the use of glass, metal, and bone as substitute materials, studies over the past 50 years have focused on the use of polymers. As the field of polymer chemistry and the understanding of its medical applications has grown, there has been a shift from the use of nonporous, rigid polymers such as polymethyl methacrylate (PMMA) to more pliable polymers, such as collagen, polyurethanes, poly(2-hydroxyethylmethacrylate) polyglycerolmethacrylate, polyvinyl alcohol, polyethylene glycol, polymethacrylic acid, silicones, polyfluorocarbons and co-polymers thereof. See U.S. Pat. No. 4,865,601; J. T. Jacob-LaBarre et al., xe2x80x9cDevelopment of a new type of artificial cornea for treatment of endstage corneal diseases,xe2x80x9d Progress in Biomedical Polymers, C G Gebelein and R L Dunn (eds.), Plenum Press, New York (1990); V. Trinkaus-Randall et al., xe2x80x9cDevelopment of a biopolymeric keratoprosthetic material,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 29(3), pp. 393-400 (1988); G. J. Crawford et al., xe2x80x9cPreliminary evaluation of a hydrogel core-and-skirt keratoprosthesis in the rabbit cornea,xe2x80x9d J. Refract. Surg., vol. 12, pp. 525-529 (1996); D. Dupont et al., xe2x80x9cBiocompatibility of human collagen type IV intracorneal implants,xe2x80x9d Cornea, vol. 8(4), pp. 251-258 (1989); W. H. Beekhuis et al., xe2x80x9cComplications of hydrogel intracorneal lenses in monkeys,xe2x80x9d Arch. Ophthalmol., vol. 105, pp. 116-122 (1987).
Recently there have been reports of success in both corneal augmentation and replacement. For replacement devices, the use of a soft, porous polymer attachment skirt has improved anchoring of the central optic and decreased extrusion of the devices. See J. T. Jacob-LaBarre et al., xe2x80x9cDevelopment of a practical artificial cornea for end stage corneal diseases,xe2x80x9d Transactions. Thirteenth Annual Meeting for the Society for Biomaterials, New York City, Volume X, p. 57 (1987); J. T. Jacob-LaBarre et al., xe2x80x9cDevelopment of a soft artificial cornea for end stage corneal diseases,xe2x80x9d Proceedings of the American Chemical Society Division of Polymeric Materials: Science and Engineering, Los Angeles, Volume 59, pp 95-99 (1988); and C. J. Crawford et al., xe2x80x9cTissue interaction with hydrogel sponges implanted in the rabbit cornea,xe2x80x9d Cornea, vol. 12(4), pp. 348-357 (1993). For augmentation devices, the use of hydrogels that allow nutrients to permeate the material to supply the tissue anterior to the implant have significantly increased the functional life of the devices. See K. P. Thompson et al., xe2x80x9cCurrent status of synthetic epikeratoplasty,xe2x80x9d Refract. Corneal Surg., vol. 7, pp. 240-247 (1991). Both types of devices share the requirement that corneal epithelial cells be able to form a viable confluent layer across the anterior surface of the optical material. However, success in that area has been limited. Although a few replacement materials have shown short-term success in maintaining a somewhat normal epithelial layer (basically a tri-cellular multilayer sheet), long-term analysis (up to 8 years) has shown that the density of the epithelial layer is not maintained. See R. A. Parks et al., xe2x80x9cHydrogel keratophakia: Long-term morphology in the monkey model,xe2x80x9d CLAO J., vol. 17(3), pp. 216-222 (1991); and T. P. Werblin et al., xe2x80x9cSynthetic keratophakia for the correction of aphakia,xe2x80x9d Ophthalmology, vol. 94, pp. 926-934 (1987).
Polymers in Biological Environments
When a synthetic material is placed in a biological environment, its surface interacts with the surrounding biological molecules. The proteins and minerals contained in the contacting interstitial fluids, such as synovial, lymphatic, or extracellular fluids, as well as blood and tears, can react specifically and non-specifically with the material surface. Over the past 30 years, extensive efforts have been made to determine the factors influencing these reactions. On a macro level, the physical characteristics of the material in relation to those of the surrounding environment (e.g., rigidity versus softness) play a critical role. Also, surface characteristics such as roughness versus smoothness influence not only fluid flow patterns and shear rates, but also the adherence of clots and cells. See S. D. Bruck, xe2x80x9cPhysiocochemical aspects of the blood compatibility of polymer surfaces,xe2x80x9d J. Polymer Sci., Polymer Symp., vol. 66, pp. 283 (1979). On a micro level, the degree of ionization and solvation of the surface and solutes and the concentration of the solutes in the surrounding media are important. The interfacial surface energy of the material also plays an important role in the thermodynamic free energy of adhesion for solutes in the surrounding media in the absence of any specific biochemical interactions (i.e., ligand-receptor interactions). See M. N. Helmus, xe2x80x9cThe effect of surface charge on arterial thrombosis,xe2x80x9d J. Biomed. Mater. Res., vol. 18, p. 165 (1984); and D. J. Lymann et al., xe2x80x9cThe effect of chemical structure and surface properties of polymers on the coagulation of blood. I. Surface free energy effects,xe2x80x9d Transact. Am. Soc. Artific. Intern. Organs., vol. XI, pp. 301-306 (1965).
Generally, proteins are intrinsically surface-active and tend to concentrate at interfaces, in part because of their polymeric structure and in part because of their typically amphoteric nature. The opportunity for multiple modes of binding with many different types of surfaces is provided by the polar, charged, and nonpolar amino acid side chains of the proteins. It has been observed that the general tendency for nonpolar residues to be internalized in the native protein often causes structural alterations of the protein upon adsorption, to maximize the number of contacts with the surface. See A. S. Hoffman, xe2x80x9cBlood-biomaterial interactions: an overview,xe2x80x9d In: Biomaterials: Interfacial Phenomena and Applications, (S. L. Cooper and N. A. Peppas (eds)), Advances in Chemistry Series 199, p. 3 (1982). Real-time FTIR analysis of proteins binding to surfaces has demonstrated conformational changes in the proteins as they first adhere and then adsorb to the material surface. See E. J. Castillo et al., xe2x80x9cProtein adsorption on hydrogels. II. Reversible and irreversible interactions between lysozyme and soft contact lens surfaces,xe2x80x9d Biomaterials, vol. 6, pp. 338-344 (1985); and E. J. Castillo et al., xe2x80x9cCharacterization of protein adsorption on soft contact lenses. I. Conformational changes of adsorbed human serum albumin,xe2x80x9d Biomaterials, vol. 5, pp. 319-325 (1984).
There are at least two different processes for the deposition and binding of proteins on surfaces: non-selective adsorption and the Vroman effect. Non-selective adsorption is based on initial polar interactions between the surface and the solutes. Specifically, for a hydrophilic hydrogel surface in a physiological medium (e.g., a soft contact lens in the tear fluid), metal ions in the hydrating fluid, such as Ca2+ and Mg2+, first interact and bind to the hydrogel matrix. The initial binding of proteins and lipids to the surface then occurs either through the polar side-chain groups of proteins or the polar head groups of the lipids, with the metal cations and hydrogel surface pendant groups acting as chelation or nucleation sites for the polar protein and lipid binding. Body beat, conformational changes, and the natural dehydration forces of hydrogen bond formation denature the proteins and lipids bound to the surface, resulting in a very thin layer of strongly bound and conformationally altered protein at the interfacial surface. Additional proteins and lipids arriving at the surface then bond with the exposed groups of the bound proteins and lipids. The new proteins are weakly bound to the first surface layer of proteins, with minimal conformational changes. Subsequent proteins and lipids become loosely associated with the weakly bound proteins, with no conformational changes. The proteins in the top layers of the adsorbed protein interact only with each other, never xe2x80x9cseeingxe2x80x9d the hydrogel surface. See Castillo, 1985; and J. L. Bohnert et al., xe2x80x9cAdsorption of proteins from artificial tear solutions to contact lens materials,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 29, pp. 362-373 (1988).
The Vroman effect proposes that protein-surface interactions in vivo (or in vitro from mixed solutions) occur as a succession of events. Molecules with low molecular weight and in high concentrations arrive first. Later, high molecular weight proteins in low concentrations displace the early arrivals. Even after the total surface concentration of protein reaches a steady state, the composition of the protein film may continue to change, with later arriving proteins eventually dominating the composition of the protein film. In this way, the surface may selectively adsorb a specific protein type or conformation. See F. C. Wedler et al., xe2x80x9cSoft contact lenses: formation of deposits,xe2x80x9d In: Biocompatibility in Clinical Practice, Vol 2, Williams D F (ed). CRC Press, Inc., Boca Raton, Fla. (1982); and C. A. LeDuc et al., xe2x80x9cA mathematical model for the Vroman effect,xe2x80x9d Ind. Eng. Chem. Res., vol. 34, pp. 3488-3495 (1995).
Once the adsorbed protein layer forms, the interactions of the biomaterial with the biological environment are mediated through this proteinaceous layer. The composition of the protein layer and the conformation and binding strength of the adsorbed proteins depend on both the surface properties of the material and the properties of the proteins present in the biological environment. See T. A. Horbett, xe2x80x9cThe role of adsorbed proteins in animal cell adhesion,xe2x80x9d Colloids and Surfaces B: Biointerfaces, vol. 2, pp. 225-240(1994). Different biological responses occur with different synthetic materials because the adsorbed protein layer can be different in each case, even when the biological environment is the same. Therefore, cells approaching the surface of a synthetic material initiate a foreign body reaction mainly in response to the characteristics of the adsorbed protein layer, unless they recognize the protein layer as xe2x80x9cself.xe2x80x9d
Recognition of the protein layer as xe2x80x9cselfxe2x80x9d is mediated by complement factors in the interstitial fluids. Although the specific sequence of protein adherence to different polymer surfaces in vivo has not been elucidated, activation of complement component C3b at the material surface has been identified.
See C. P. Klein et al., xe2x80x9cActivation of complement C3 by different calcium phosphate powders,xe2x80x9d Biomaterials, vol. 4, pp. 181-184 (1986). Once activated, C3b (either surface-bound or free) can interact directly with macrophages and indirectly with T cells by several mechanisms to initiate a foreign body response. Bound C3b can mediate macrophage attachment to the material""s surface. (Macrophage adhesion leads to the release of lysosomal enzymes and other cytotoxic components from the macrophages.) The intensity or degree to which the foreign body response is maintained at the material""s surface determines the ability of healthy cells to proliferate and spread on or near that surface. See A. Remes et al., xe2x80x9cImmune response in biocompatability,xe2x80x9d Biomaterials, vol. 13(11), pp. 731-743 (1992).
Cell Adhesion to Polymers
Cell adhesion to a material, whether that material is the naturally occurring extracellular matrix (ECM) of the basement membrane or a synthetic biomaterial, is mediated primarily by the interaction between surface-bound proteins and the corresponding receptors on the cell membrane. Many years of biomaterial research have focused on optimizing the material surface characteristics, such as charge, energy, and roughness, to influence specific protein deposition, to try to tailor the Vroman effect for the specific type of cellular adhesion desired. Studies involving both fibroblasts and endothelial cells in culture have shown that positively charged surfaces enhance cell proliferation and adhesion significantly better than negatively charged or non-ionic hydrogel surfaces. See S. Hattori et al., xe2x80x9cFibroblast cell proliferation on charged hydroxyethyl methacrylate copolymers,xe2x80x9d J. Colloid Interface Sci., vol. 104, pp. 72-78 (1985); and X. Y. Wu et al., xe2x80x9cExpression of integrin and organization of F-actin in epithelial cells depends on the underlying surface,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 35, pp. 878-890 (1994). High surface energy or wettability has been shown to interfere with both human fibroblast and endothelial cell attachment, and low surface energy or wettability has been shown to interfere with plasma protein adsorption. See H. Yasuda et al., xe2x80x9cThe rate of adhesion of melanoma cells onto nonionic polymer surfaces,xe2x80x9d J. Biomed. Mater. Res., vol. 12, pp. 701-706 (1978); and G. Altankov et al., xe2x80x9cReorganization of substratum on hydrophilic and hydrophobic materials is related to biocompatibility,xe2x80x9d J. Mater. Sci., vol. 5, pp. 732-737 (1994). In general, non-wettable materials have been shown to inhibit the attachment and growth of anchorage-dependent cells. See T. Groth et al., xe2x80x9cStudies on cell-biomaterial interaction: role of tyrosine phosphorylation during fibroblast spreading on surfaces varying in wettability,xe2x80x9d Biomaterials, vol. 17, pp. 1227-1234 (1996). Recent studies have shown that microscopically roughened surfaces demonstrate enhanced vascular cell adhesion and migration rates, compared with smooth surfaces of the same material. See M. Lampin et al., xe2x80x9cCorrelation between substratum roughness and wettability, cell adhesion, and cell migration,xe2x80x9d J. Biomed. Mater. Res., vol. 36, pp. 99-108 (1997); and S. L. Goodman et al., xe2x80x9cThree-dimensional extracellular matrix textured biomaterials,xe2x80x9d Biomaterials, vol. 17, pp. 2087-2095 (1996).
Further research has indicated the importance of the conformation of the protein in promoting interaction with cells. Studies have shown that extracellular matrix proteins such as fibronectin and vitronectin must be adsorbed from the plasma or serum onto biomaterial surfaces as a prerequisite for successful cell adhesion and spreading. See J. G. Steele et al., xe2x80x9cAttachment of human bone cells to tissue culture polystyrene: the effect of surface chemistry upon initial cell attachment,xe2x80x9d J. Biomater. Sci. Polymer Edn., vol. 5, pp. 245-257 (1993). However, adsorption of fibronectin onto hydrophobic materials is followed by a decrease in cell adhesion, adhesion strength, and diminished cell spreading, a fact that has been attributed to the possible conformation changes in the protein induced upon adsorption. See D. J. Juliano et al., xe2x80x9cEffect of the conformation and orientation of adsorbed fibronectin on endothelial cell spreading and the strength of adhesion,xe2x80x9d J. Biomed. Mater. Res., vol. 27, pp. 1103-1113 (1993); and J. D. Andrade et al., xe2x80x9cProtein adsorption and materials biocompatibility: A tutorial review and suggested hypothesis,xe2x80x9d Prog. Surface Sci., vol. 79, pp. 1-64 (1986). This phenomenon spurred investigators to pre-adsorb different proteins onto surfaces and determine the cellular response. Collagen, laminin, vitronectin, fibronectin, and albumin have been pre-adsorbed onto a variety of surfaces in an attempt to find appropriate binding substrates. See R. Z. Xie et al., xe2x80x9cEffects of biologically modified surfaces of synthetic lenticules on corneal epithelialization in vivo,xe2x80x9d Austl. N.Z.J. Ophthalmol., vol. 25(Suppl 1), pp. S46-S49 (1997); S. M. Kirkham et al., xe2x80x9cThe keratoprosthesis: improved biocompatibility through design and surface modification,xe2x80x9d Ophthalmic Surg., vol. 22, pp. 455-461 (1991); V. Trinkaus-Randall et al., xe2x80x9cModification of polymers for synthesis by corneal epithelial cells,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 32(suppl), pp. 1072 (1991); and J. G. Steele et al., xe2x80x9cMechanism of initial attachment of corneal epithelial cells to polymeric surfaces,xe2x80x9d Biomaterials, vol. 18, pp. 1541-1551 (1997). Additionally, increased amounts of absorbed fibronectin are accompanied by decreased recognition of the cell-binding domain. See D. K. Pettit et al., xe2x80x9cCorrelation between corneal epithelial cell outgrowth and monoclonal antibody binding to the cell binding domain of adsorbed fibronectin,xe2x80x9d J. Biomed. Mater. Res., vol. 28, pp. 685-691 (1994). It has also been reported that cells are able to reorganize and rearrange adsorbed fibronectin on a hydrophilic glass surface but not on a hydrophobic surface. Cells can also increase the concentration of fibronectin on the surface beneath the cells. See G. Altankov et at, xe2x80x9cReorganization of substratum-bound fibronectin on hydrophilic and hydrophobic materials is related to biocompatibility,xe2x80x9d J. Mater. Sci. Mater. Med., vol. 5, pp. 732-737 (1994). Tyrosine phosphorylation in fibroblasts on surfaces of varying wettability has been examined, and the studies indicated that the adverse effect of hydrophobic surfaces on cell adhesion and proliferation was due to an impaired transfer of signals via integrins from the substratum (adsorbed fibronectin) to the cell interior. See T. Groth et al., xe2x80x9cStudies on cell-biomaterial interaction: role of tyrosine phosphorylation during fibroblast spreading on surfaces varying in wettability,xe2x80x9d Biomaterials, vol. 17, pp. 1227-1234 (1996).
Tissued Engineered Surfaces
To improve cell adhesion and spreading, short peptide sequences, e.g., YIGSR and RGD, that are responsible for cell-surface adhesion binding activity in extracellular adhesion proteins (fibronectin and laminin) have been chemically incorporated onto polymer surfaces. See S. P. Massia et al., xe2x80x9cCovalent surface immobilization of Arg-Gly-Asp- and Tyr-Ile-Gly-Ser-Arg-containing peptides to obtain well-defined cell-adhesive substrates,xe2x80x9d Analyt. Biochem., vol. 187, pp. 292-301 (1990); S. P. Massia et al., xe2x80x9cCovalently immobilized laminin peptide Tyr-Ile-Gly-Ser-Arg (YIGSR) supports cell spreading and co-localization of the 67-kilodalton laminin receptor with xcex1-actinin and vinculin,xe2x80x9d J. Biol. Chem., vol. 268, pp. 8053-8059 (1993); H.-B. Lin et al, xe2x80x9cSynthesis, surface, and cell-adhesion properties of polyurethanes containing covalently grafted RGD-peptides,xe2x80x9d J. Biomed. Mater. Res., vol. 28, pp. 329-342 (1994); and J. R. Glass et al., xe2x80x9cCharacterization of a hyaluronic acid-Arg-Gly-Asp peptide cell attachment matrix,xe2x80x9d Biomaterials, vol. 17, pp. 1101-1108 (1996). Although these minimal binding sequences have only a fraction of the activity of the entire protein, their small size allows them to be incorporated at much higher concentrations than would be possible with entire proteins. The short peptide sequences have the advantage of being relatively stable, and their synthetic nature renders them amenable to chemical derivatization and covalent attachment.
Direct attachment of these cell-surface receptor recognition sequences has been associated with an increase in cell culture adhesion for a variety of cell types, including human foreskin fibroblasts, bovine pulmonary endothelial cells, human umbilical vein endothelial cells, and porcine pulmonary aortic endothelial cells. See the references cited in the preceding paragraph. However, the increase in cell adhesion response in these systems, while significant, was not as high as had been hoped, and the adhesion levels did not reach those seen in natural systems. Generally, the increase seemed to be of a nonspecific nature and was not found for all cell types tested. See D. L. Hern et al., xe2x80x9cIncorporation of adhesion peptides into nonadhesive hydrogels useful for tissue resurfacing,xe2x80x9d J. Biomed. Mater. Res., vol. 39, pp. 266-276 (1998). Use of these sequences on surfaces cultured with rabbit corneal epithelial cells has not demonstrated the same marked increase in cell adhesion as has been reported for other cellular systems. See V. Trinkhaus-Randall et al., 1991; and K. Watanabe et al., xe2x80x9cMechanisms of persistent epithelial defect formation: peptide from the cell-domain (GRGDS) inhibits corneal epithelial cell migration,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 29(suppl), p. 192 (1988). More recently, the fibronectin adhesion-promoting peptide sequence, H-trp-gln-pro-pro-arg-ala-arg-ile-OH, (FAP), a peptide encompassing the heparin binding domain of the fibronectin molecule has shown some ability to increase corneal epithelial cell proliferation. See D. L. Mooradian et al, xe2x80x9cCharacterization of FN-C/H-V, a novel synthetic peptide from fibronectin that promotes rabbit corneal epithelial cell adhesion, spreading and motility,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 34, pp. 153-164 (1993).
Other research has shown that attaching a peptide sequence to a spacer arm resulted in a 50% increase in specific cell attachment, compared with the response to a surface with directly attached peptides. See K. Kugo et al., xe2x80x9cFibroblast attachment to Arg-Gly-Asp peptide-immobilized poly(gamma-methyl L-Glutamate),xe2x80x9d J. Biomater. Sci., Polymer Edn., vol. 5, pp. 325-327 (1994); and D. L. Hern et al., 1998. However, other research has indicated that cell adhesion proteins have multiple attachment sites. Five peptide sequences on the laminin alpha 1 chain carboxyl-terminal globular domain have been found to exhibit cell-type-specific attachment activities, including SIYITRF, IAFQRN, and LQVQLSIR. See M. Nomizu et al., xe2x80x9cIdentification of cell binding sites in the laminin alpha 1 chain carboxyl-terminal globular domain by systemic screening of synthetic peptides,xe2x80x9d J. Biol. Chem., vol. 270, pp. 20583-20590 (1995). PHSRN, a synergistic peptide that enhances the activity of the RGD peptide, has been found in the central cell-adhesive domain of fibronectin. See S. Aota et al., xe2x80x9cThe short amino acid sequence Pro-His-Ser-Arg-Asn in human fibronectin enhances cell-adhesive function,xe2x80x9d J. Biol. Chem., vol. 269, pp. 24756-24761 (1994). Additionally, in rabbit corneal epithelial cells an interaction between cells and fibronectin via multiple adhesion-promoting sequences within the intact fibronectin molecule was shown. See J. D. Cameron et al., xe2x80x9cRabbit corneal epithelial cell adhesion to proteolytic fragments and synthetic peptides of fibronectin,xe2x80x9d Invest. Ophthalmol. Vis. Sci., vol. 32(suppl), pp. 1072 (1991).
Recently investigators have reported the use of tethering biologically active molecules to polymer scaffolds for tissue regeneration. See L. Griffith Cima, xe2x80x9cPolymer substrates for controlled biological interactions,xe2x80x9d J. Cell. Biochem., vol. 56, pp. 155-161 (1994); and P. R. Kuhl et al., xe2x80x9cTethered epidermal growth factor as a paradigm for growth factor-induced stimulation from the solid phase,xe2x80x9d Nature Med., vol. 2, pp. 1022-1027 (1996). By covalently linking epithelial growth factor (EGF) onto a star-poly(ethylene oxide) (PEO) tether and then anchoring the tether onto the surface of a biodegradable scaffold, a 40% increase in rat hepatocyte cell adhesion and migration was shown. See L. Griffith-Cima, xe2x80x9cTissue engineered scaffolds for liver regeneration,xe2x80x9d Presented at Molecular Engineering of Polymers workshop: Directing Biological Response, American Chemical Society, November, 1996. Moreover, it was also shown that DNA synthesis within the cells was comparable to the levels found when the medium contained free EGF.
U.S. Pat. No. 5,986,043 discloses photopolymerizable biodegradable hydrogels for use in reducing the formation of cell adhesion after surgery, in applying a drug locally to a tissue surface, and in adhering tissue surfaces in a patient.
U.S. Pat. No. 5,906,828 discloses growth effector molecules, including growth factors and extracellular matrix molecules, flexibly linked by branched tethers to a support medium; and the use of the combination to stimulate and support cell and tissue growth.
U.S. Pat. No. 5,836,313 discloses a two-layer composite material composed of a thin-layer tissue and a hydrogel which is designed to provide a suitable substrate for corneal epithelial cell growth while maintaining the clarity, flexibility, and diffusivity of hydrogels.
U.S. Pat. Nos. 5,830,504 and 5,654,267 disclose a composition of an xcex1xcex3xcex23 integrin ligand and a growth factor receptor ligand combined in a matrix that was said to be useful for promoting wound healing and tissue regeneration.
U.S. Pat. Nos. 5,760,176 and 5,120,829 disclose a method to attach a peptide to a solid substrate using its hydrophobic domains.
U.S. Pat. No. 5,716,633 discloses a collagen-hydrogel fabricated into an artificial lens which is capable of promoting epithelial cell growth.
U.S. Pat. No. 5,677,276 discloses peptides conjugated to hyaluronate polymers which may be used to promote the healing of wounds and tissue regeneration.
U.S. Pat. No. 5,512,474 discloses a cell culture system which comprises a support material with a surface bearing a combination of a positively-charged molecule and a cell adhesion factor.
U.S. Pat. No. 5,278,063 discloses a method to chemically graft peptides to a surface to enhance cell-surface adhesion to optimize cell culture systems and to improve cell bioadhesion to surfaces made of various materials.
U.S. Pat. No. 5,171,264 discloses hydrogels produced by covalently immobilizing polyethylene oxide star molecules onto a support surface.
Attempts to augment and replace the cornea with synthetic materials that preferably have a refractive index close to that of the normal cornea have met with limited success primarily because these devices have been unable to support and maintain a normal stratified epithelium as an effective barrier. The fundamental problem is that the host epithelial cells recognize the polymer surface as other than the normal stroma layer or xe2x80x9cself,xe2x80x9d triggering an immunological response throughout the life of the implant. Although some surfaces have shown a degree of epithelial growth, the resulting epithelium has been irregular, lacking the normal tri-layer appearance.
We have developed a synthetic device for cornea augmentation and replacement that increases corneal epithelium cell adhesion and migration. Additionally, the device can be used to improve vision. By using tethered corneal enhancer molecules, for example, extracellular matrix proteins (ECMPs), corneal growth factors, and other ligand-specific enhancer molecules on the surface of an optical polymer for use as corneal device, the epithelial cell response can be significantly enhanced for the following reasons: a) the tethers (or chains) hold the corneal enhancer molecules to the surface in a favorable conformational state; and b) the tethers hold the corneal enhancer molecules far enough away from the surface that they are not obscured by non-specific protein binding to the surface. By tethering a combination of corneal enhancer molecules, a more natural environment can be created. Additionally, the surface topography of the artificial surface, preferably a hydrogel, can be micro-molded, lathed, or engineered prior to tethering the corneal enhancer molecules to resemble the natural underlying surface of the corneal epithelial cells, Bowman""s layer. This system allows epithelial cells to spread and attach faster than existing systems, as well as providing an underlying textured surface that allows the cells to resist the shear force induced in vivo by the blinking of the eyelid. Moreover, the resulting epithelial layer closely resembles a natural epithelial layer. The material can be used, for example, as a corneal onlay, an epikeratophakia lenticule, an intracorneal augmentation device, or an artificial cornea.
As used in the specification and claims, xe2x80x9coptical polymerxe2x80x9d refers to a device or part of a device which can be manufactured to be used to augment or replace the natural cornea. Preferably the optical polymer has a refractive index close to that of a natural cornea. Examples of polymers useful in making an optical polymer include collagen, polyurethanes, poly(2-hydroxyethylmethacrylate), polyglycerolmethacrylate, polyvinylpyrolidone, polyvinyl alcohol, polyethylene glycol, polymethacrylic acid, silicones, polyfluorocarbons, polymers with phosphocholine groups and co-polymers or combinations thereof. Some of these polymers will form a surface with reactive chemical species to which a tether can attach. Other polymers will need to have the surface treated by methods known in the art to produce reactive species for linkage of the tether, for example, by plasma etching, ion beam, corona discharge, electron beam, ion exchange, UV irradiation, or gamma irradiation.
As used in the specification and claims, xe2x80x9ctetherxe2x80x9d refers to a polymer or chain of chemical molecules that serves as a link between the surface of the optical polymer and a corneal enhancer molecule. The tether is preferably a linear, single chain polymer for example, such as polyethylene oxide (PEO), without any significant pendant groups. The linear, single chain tether may also be made from amino acids or peptides. The molecular weight of the tether is preferably from about 2000 to about 8000, and most preferably about 3400. The tethers are covalently linked to the surface of the optical polymer and while these bonds are preferably hydrolyzable, they can be non-hydrolyzable in nature. The form depends on the tether material and the surface of the optical polymer.
As used in the specification and claims, xe2x80x9ccorneal enhancer moleculexe2x80x9d refers to a biological molecule that binds to a corneal epithelial cell receptor and promotes the growth and expansion of the corneal epithelial cell layer. Corneal enhancer molecules include extracellular matrix proteins, corneal growth factors, and other ligand-specific corneal enhancer species that promote corneal epithelium growth and migration. Examples of extracellular matrix proteins include fibronectin, laminin, kalinin, K-laminin, vitronectin, talin, integrin, and albumin. Examples of corneal growth factors include insulin-like growth factor (IGF), fibroblast growth factor (FGF), hepatocyte growth factor, epithelial growth factor (EGF), transforming growth factor-xcex1 (TGF-xcex1), transforming growth factor-xcex2 (TGF-xcex2), keratinocyte growth factor (KGF), heparin binding factor, and nerve growth factor. Examples of other ligand-specific corneal enhancer species include the neurotransmitter, substance P; the cytokines, interleukin-1 alpha, and interleukin-1 beta; and peptides, including fibronectin adhesion-promoting peptide sequence consisting of H-trp-gln-pro-pro-arg-ala-arg-ile-OH, YIGSR, SIYITRF, PHSRN, IAFQRN, and LQVQLSIR.
The effects of tethered corneal enhancer molecules on rabbit epithelial cell proliferation and adhesion have been measured. Tethers of 3400 MW poly(ethylene glycol) (PEG) (also known as polyethylene oxide, PEO) have been used to attach specific factors and proteins to a hydrogel surface. PEG does not interact with most macromolecules found in body fluids. For example, see K. Bergstrom et al., xe2x80x9cReduction of fibrinogen adsorption on PEG-coated polystyrene surfaces,xe2x80x9d J. Biomed. Mater. Res., vol. 26, pp. 779-790 (1992). Tethering of the active molecules decreases unproductive diffusion and increases the probability of successful interactions with target molecules or cells. Additionally, we have determined that if the linear tether is of sufficient length (about 30-150 nm), it kept the tethered-molecule away from the hydrogel surface and removed from any non-specific protein that may absorb to the hydrogel surface. From prior contact lens research, it is known that the adsorbed protein layer on hydrogel lenses is generally in the range of 0.6 to 1xc2x0 g/cm2 in a tri-layer configuration, corresponding to a thickness that is less than the effective length of the tether. Proteins passively adsorbed to the surface will therefore not interfere with the function of the tethered molecules in enhancing epithelial cell attachment to the tether-modified hydrogel surfaces. Although the linear tether is mobile, and can twist and bend, it is generally thermodynamically unfavorable for the tether to bend sufficiently to allow the tethered molecule to interact with adsorbed protein on the hydrogel surface. Unlike a branched or star formation tether, a linear chain allows the molecule attached to its free end to rotate and move in space. The linear tether allows the molecule to orient itself in the most favorable conformation for binding to the cell membrane.
By tethering laminin, fibronectin, substance P, insulin-like growth factor-1 (IGF-1) and ligand-specific chemical molecules such as the peptides that make up the carboxy-terminal heparin binding and cell adhesion-promoting domain of fibronectin [FAP], either singly or in various combinations, the natural cell adhesion and proliferation on a polymer surface is increased. The maintenance of the normal integrity of the corneal epithelium is regulated by both humoral and neural factors. Tether-modification involving various corneal enhancer molecules, such as laminin, fibronectin, substance P, FAP and IGF-1, was used to optimize the effects of these molecules on epithelial cell adhesion and spreading on synthetic biomaterial surfaces for uses in augmenting and/or replacing portions of the natural cornea. Additionally, a combination of laminin and fibronectin in approximately equal molar amounts was used. Adhesion of cells to a biomaterial surface is a major factor affecting biocompatibility. The addition of biological molecules to biomaterial surfaces should directly affect cell adhesion strength. The optimal tethered molecule-modified surfaces would have adhesion strengths more similar to those of natural stroma than those of protein-coated surfaces. This indicates that epithelial cells interact with the tethered molecule-modified surfaces as strongly as they interact with natural stroma.
Due to the specific radius of movement of the tether-molecule complex on the hydrogel surface, there is an optimal density of tethered proteins that will produce the maximal epithelial cell spreading and attachment response. By tailoring the number of binding sites on the polymer surface during the polymerization and by knowing the efficiency of binding, tethered-modified polymers may be optimized with differing densities of tethered-molecules on the surfaces.
Additionally, the surface topography of the synthetic corneal surface may play an important role in increasing and maintaining cell attachment in vivo. Although the exact function of Bowman""s layer is not known, the micro and/or nanoscale texture imparted by this anterior layer of the stroma may be important in allowing the epithelial cells to withstand the shear force induced by blinking of the eyelids. The anterior surface of the synthetic material can be designed to mimic the surface topography of Bowman""s layer. Several techniques may be used to produce the anterior surface such as micro-molding, etching, lathing or polymer engineering. To the resultant three-dimensional surface, corneal enhancer molecules of interest can be tethered. Such an engineered synthetic material which can support a normal corneal epithelium can be used for a corneal onlay, an epikeratophakia lenticule, an intracorneal augmentation device, or an artificial cornea.